基于SM-PEPA可生存系统认知模型及量化分析  

Cognitive model and quantitative analysis for survivable system based on SM-PEPA

作　　者：王健[1] 赵国生[2] 

机构地区：[1]哈尔滨理工大学计算机科学与技术学院,黑龙江哈尔滨150080 [2]哈尔滨师范大学计算机科学与信息工程学院,黑龙江哈尔滨150025

出　　处：《华中科技大学学报（自然科学版）》2015年第5期99-103,共5页Journal of Huazhong University of Science and Technology(Natural Science Edition)

基　　金：国家自然科学基金资助项目(61403109;61202458);高等学校博士学科点专项科研基金资助项目(20112303120007)

摘　　要：为解决可生存系统的自主管理和维护问题以保持关键服务的持续提供,提出一种基于SM-PEPA的系统认知形式化模型及量化分析方法.首先建立了具有分层认知能力的可生存系统认知模型,即服务认知子层、网络认知子层和接入认知子层,然后提出了一种由策略库发起驱动基于认知环MDE的认知单元自管理模型,并通过可生存系统的生存状态转换图,构建了可生存系统认知性能分析框架.在此基础上,结合对可生存系统认知能力的形式化描述,将其转化为一个半马尔可夫过程,对模型进行了量化分析.仿真试验分析了不同参数变化对系统可生存性的影响,试验结果验证了所提方法的有效性和合理性.As self-managing,maintaining and keeping the key services of survivable systems is becoming more and more difficult,a method of cognitive formalized model and quantitative analysis for survivable system was proposed based on semi-Markov performance evaluation process algebra（SMPEPA）.Firstly,a cognitive model of survivable system with hierarchical cognitive ability was established,namely service cognitive sublayer,network cognitive sublayer and access cognitive sublayer.Then,a self-management cognitive unit model derived by strategy library based on monitor-decide-execute was put forward.Finally,aperformance analysis framework of cognitive model was built according to the transferring graph of system state for survivability.On the basis,the quantitative analysis was carried out,with the formal description of cognitive ability considered and which was transformed to a semi-markov process simultaneously.Simulation experiments discusses the effect of survivability according to different parameters,and the experimental results show the effectiveness and feasibility of the developed method.

关 键 词：认知模型 随机进程代数 可生存性 反馈控制环 量化分析 

分 类 号：TP302[自动化与计算机技术—计算机系统结构]